Economy & Finance

Determinants of Foreign Direct Investment (FDI) in Malaysia: What Matters Most?

1. International Review of Business Research Papers Volume 6. Number 6. December 2010 Pp.235 –245 Determinants of Foreign Direct Investment in Malaysia: What Matters…
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  • 1. International Review of Business Research Papers Volume 6. Number 6. December 2010 Pp.235 –245 Determinants of Foreign Direct Investment in Malaysia: What Matters Most? Nursuhaili Shahrudin *, Zarinah Yusof** and NurulHuda Mohd. Satar*** This paper examines the determinants of foreign direct investment in Malaysia for the period 1970-2008. The causality and dynamic relationships between foreign direct investment (FDI) and its key determinants is identified using autoregressive distributed lag (ARDL) framework. The result suggests that among the variables, financial development and economic growth contribute positively to the inflow of foreign direct investment in Malaysia. The establishment of more financial intermediaries with a modern and sophisticated technology in the banking system creates a favorable environment to the foreign investors. The evidence provides strong policy recommendation on sustaining high growth and financial deepening in Malaysia.Field of Research: Macroeconomics1. IntroductionMalaysia is one of the countries in Asia that has benefited from strong foreigndirect investment inflow. FDI was a major source of growth for manufacturingdevelopment in Malaysia that mainly targeted for the export market. Theeconomy relied on the foreign fund as a major source of capital, moderntechnology and technical skills. Globalization, international financialintegration and expansion of global production have intensified FDI in the pastdecades. The inflow of FDI was only US$94 million in the early 1970s butincreased dramatically to US$7,297 million in 1996. The amount dropped toUS$2,714 million in 1998 due to the 1997 financial crisis but recoveredstrongly with inflows of US$8,403 million in 2003. Malaysia was among thetop attractive FDI destination to many investors in Southeast Asian countries.Last year, FDI was about US$7.3 billion. On another front, Malaysia wasmore bullish attracting FDI from the ASEAN region, China, India, South Koreaand the Middle East. Most notably the environment investment has becomingmore challenging and competitive for FDI. This paper identifies some of themain factors that contribute significantly to attracting FDI in Malaysia. This iscrucial because private investment has been given a leading role to bring theeconomy to higher growth and sustainable economy._______________*Nursuhaili Shahrudin, Faculty of Economics and Administration, University Malayaemail:**Dr. Zarinah Yusof, Faculty of Economics and Administration, University Malaya email:***Dr. NurulHuda Mohd. Satar, Faculty of Economics and Administration, University Malaya email:
  • 2. Shahrudin, Yusof & Satar2. Literature ReviewFDI is motivated mainly by the possibility of high profitability in growingmarkets, along with the possibility of financing these investments at relativelylow rates of interest in the host country. The basic concept of internationaltrade theory was founded by David Ricardo arguing that countries engaged inforeign trade due to comparative cost advantages. Heckscher, Ohlin, andSamuelson further developed Ricardo’s theory of comparative advantage byarguing that a country specialized in producing and exporting goods in whichit has comparative advantage due to natural endowments of the factors ofproduction in the host country relative to elsewhere. This suggests that theability to produce at lower opportunity cost relative to another country is amajor factor in the location decision (Ohlin, 1933).Recently, the Gravity Model has become a popular method to analyze theimportance of countries’ attractive location factors for FDI. The Gravity Modelis based on the interactions of various potential sources across border. Themodel has an inverse relationship between country of origin and destinations.The basic gravity model proposes that trade between two countries is afunction of the size of their economies as measured by the national incomeand population and the geographical distance between the two countries(Breuss and Egger, 1997).There are few studies on FDI and its determinants in Malaysia. Some studieson FDI and its determinants are found in Zubir Hassan (2003), Hooi (2008)and Ang (2008). Zubair Hasan (2003) shows that foreign exchange rate,export expansion, and infrastructural development are important factors inattracting FDI into Malaysia. The exchange rate variable has a negativerelationship with FDI. This implies that a weak currency reduces FDI inflowswhich contradict the hypothesized sign. Other factors such as growth rate,capital flight, and balance of payments have smaller impact on FDI inflows.The Engel-Granger test for cointegration however failed to support any long-term relationships between FDI and each explanatory variable.In the same spirit, Hooi (2008) used Granger causality test and errorcorrection method to examine the impact of FDI on growth and growth onFDI. The Indirect impact of FDI on growth is analyzed through manufacturingsector. He finds that the relationship between FDI and growth of themanufacturing sector is independent of each other and concludes that there isno long-run relationship between FDI and growth.Study by Ang (2008) includes a wider coverage of FDI determinants with dataspanning from 1960 to 2005. Among the variables tested are financialdevelopment, wage rates, income, economic growth, government spendingon infrastructure, openness, exchange rate, inflation rate, corporate tax andfinancial crisis 1997-1998. The findings suggest that all the factors areimportant determinants of FDI. Economic growth rate has the smallest effecton FDI while exchange rate has the biggest role in attracting more FDI. Thestudy concludes that higher corporate tax rate and ringgit appreciation havedampen effects on FDI inflows. Surprisingly, the study shows that 236
  • 3. Shahrudin, Yusof & Satarmacroeconomic uncertainty promotes greater FDI. The possible explanationgiven to this is investors perceived that during macroeconomic uncertaintythere is a greater potential for investment return.The highly significant causal relationships of the variables in Ang’s study raisesome doubts. Although the problem of endogeneity bias is addressed via2SLS estimation, the stationarity properties of the series are not tested. It iswidely found that most time series data are nonstationary series, hence thehigh correlation among the variables may not reflect their true relationshipsbut simply they are trending together over time. Therefore, the estimation ofthe dynamic relationships among the variables needs a different estimationtechnique. Moreover, the highly statistically significant estimates of thecoefficients are the results of five different equations that exclude some of thevariables from the equations. While the evidence of no long-run relationshipbetween FDI and its determinants in the previous studies is due to the failureto model cointegration relationship appropriately.Thus, these issues provide the motivational of this paper. This study aims toinvestigate the determinants of FDI in Malaysia for the period 1970-2008using different approach analyzing the dynamics of the variables. This modelapplies cointegration analysis based on an autoregressive distributed lag(ARDL) technique. Specifically, it seeks to identify the most importantvariables that affect the FDI inflow in Malaysia and provides some newevidence on the existence of long-run relationship between FDI and itsdeterminants.3. Methodology and Data3.1 Autoregressive distributed lag (ARDL)Many macroeconomic variables are non-stationary at their level. A linearcombination of non-stationary variables does not imply that all the variablesare cointegrated. If the series are cointegrated, an error correction modelshows that the short-run dynamics of the variables in the system areinfluenced by the deviation from equilibrium (Enders, 2004). Twodistinguished advantages of using autoregressive distributed lag model are itallows cointegration of variables with different orders of integration and it is arobust estimation for a small sample data. Failure to model appropriately therelationships may not give accurate results of the relationships and this iscrucial especially when it involves policy recommendations. The existence oflong-run relationship between FDI and selected macroeconomics variable ismodeled as follows.Lfdi = f ( Lm2, Lgdp, gro , Ldev ,open, Lexc, tax, infl, crisis )where L refers to the variable in logarithmic form, fdi is the foreign directinvestment inflows (RM million), m2 is the money supply i.e. a proxy forfinancial market development, gdp represents the market size of the economy(RM million) and gro is the growth rate of gross domestic product (%).Government infrastructure expenditure (RM million) is represented by variable 237
  • 4. Shahrudin, Yusof & Satardev, open is the ratio of import and export over GDP which measureseconomic openness, exc is the exchange rate (RM/USD), tax is the corporatetax (%), infl is the inflation rate (%) and crisis is a dummy variable representsthe effect of Asian financial crisis (1998=0 and otherwise=0). The expectedsign of Lm2, Lgdp, gro , Ldev ,open, Lexc is positive and negative for tax, infl,crisis variables.Financial development is used to measure financial depth of a country. Amore developed financial system allows an economy to exploit the benefits offoreign direct investment more efficiently (Ang, 2009). Therefore,advancement in financial market system will attract more foreign investments.On the other hand, some studies also emphasize on the importance of largemarket size for efficient utilization of resources and exploitation of economiesof scale (Zaman et al., 2006). An economy with a large market size shouldattract more FDI into the country. Market size is important for FDI as itprovides greater profitability from selling in local markets than selling to exportmarkets. Thus, a large market size provides more opportunities for sales andalso profits to MNCs and therefore attracts FDI. However, the market sizemight be less influential or insignificant if foreign companies are using thehost country only as a production base for export in a way that the MNCs aretaking advantage of lower costs of production in countries like Malaysia andChina. Many studies have proved that market size, represented by nominalGDP is significant in attracting FDI.Countries with high and sustained growth rate, generally measured by GDPgrowth, are more likely to receive larger amount of FDI compared to othercountries. There are numerous studies that show a positive relationshipbetween growth rate and FDI (Sahoo, 2006). A stable economic growth ratesignifies a good economic performance and therefore is more attractive toforeign investors. Another important factor that potentially affect foreigninvestors’ decision in choosing a host country is the availability of physicalinfrastructure such electricity, water, transportation, and telecommunications.Previous studies have used several different proxies to measure theinfrastructure development. Among them are total government spending ontransport and communication (Ang, 2008), telephones per 1,000 population(Asiedu, 2002), infrastructure index (Sahoo, 2006), and governmentdevelopment expenditure (Zubair Hasan, 2003). Due to data availability, totalexpenditure by central government is employed as it accounts for not onlytransportation and communication, electricity, and water but it also includesexpenditures on all economic sectors. Increase in development expenditureshould have a positive impact on FDI.Openness of the domestic economy is influenced both by direct FDIrestrictions as well as trade barriers. FDI restrictions clearly raise barriers toFDI and are likely to influence the choice that the MNCs make regarding theinvestment location. The ratio of trade to GDP is often used as a measure ofopenness of an economy in many literatures. The other variables that canalso measure economic openness are ratio of export to GDP and ratio ofimport to GDP. The impact of openness of an economy towards FDI woulddepend on the investment type. Trade restrictions and thus less openness of 238
  • 5. Shahrudin, Yusof & Sataran economy is associated with market-seeking investment that would affectFDI positively. This is because if MNCs face difficulties in terms of traderestrictions in the host country, they would set up subsidiaries in order toserve the host country local market. On the other hand, a more openeconomy would be an attractive location for MNCs that operate in exporting toreduce the transaction costs (Asiedu, 2002). Since the motive of MNCsengage in FDI is to export their products we expect that trade opennesswould have a positive relationship with FDI inflow.Exchange rate is an important determinant of international capital flows in thefinancial market (Hsiao & Hsiao, 2004), although its effects on FDI arecomplicated and the direction of influence on FDI is not well-established.Tahir and Larimo (2005) argued that a country can attract FDI by devaluing itscurrency because FDI will benefit from currency weakness in the hostcountry. The depreciation of foreign investors’ country currency againstMalaysia Ringgit would increase the inflows of FDI. The exchange rate is oneof the most significant factors affecting trade between countries. If theexchange rate rises, trade is relatively more profitable to exporters, soexporters will be sensitive to changes in the exchange rate. Statutorycorporate tax rate is used as a proxy for the effects of all fiscal policies onnew investors, ignoring tax holidays, accelerated depreciation and otherincentives that reduce the effects of the statutory rate. Higher corporate taxrate imposed to corporate profits reduces FDI returns and hence it wouldnegatively affect FDI inflows.In this study, the rate of inflation acts as a proxy for the level of economicuncertainty, a measure of overall economic stability. If investors prefer toinvest in more stable economies which it reflects a lesser degree of economicuncertainty, thus lower inflation would stimulate more FDI into the country.The hypothesized signs of the selected variables are as follows. Table 1: Variables Descriptions and Expected SignsVariable Description Expected SignLm2 Money supply (M2) +Lgdp Gross Domestic Product +gro Gross Domestic Product growth rate +Ldev Government Development Expenditure +open Ratio of Import and Export over Gross + Domestic ProductLexc Exchange rate (RM/USD) +tax Corporate tax -infl Inflation rate -crisis Value equal to “1” for year 1998 and “0” - otherwise 239
  • 6. Shahrudin, Yusof & Satar3.2 Estimating modelThe econometrics model of the causality of FDI and its key determinants is asfollows: Lfdit   0  1 Lm 2t   2 Lgdpt   3 grot   4 Ldevt   5 opent   6 Lexc t   7 taxt   8 inf lt   9 crisist   t (1)In testing the existence of the long-run relationship (cointegration), the errorcorrection versions of the ARDL framework for equation (1) is given byequation (2). n n n n n Lfdit   0   bi Lfdit i   ci Lm2 t i   d i Lgdp t i   ei grot i   f i Ldevt i i 1 i 1 i 1 i 1 i 1 n n n n n   g i opent i   hi Lexc t i   ji taxt i   k i  inf lt i   li crisist i i 1 i 1 i 1 i 1 i 1   1 Lfdit 1   2 Lm2 t 1   3 Lgdp t 1   4 grot 1   5 Ldevt 1   6 opent 1   7 Lexct 1   8 taxt 1   9 inf lt 1   10crisist 1   t (2)Equation (3) presents the error correction version of the ARDL frameworkwhen the variables are Ldfi, Lm2 and gro. n n n Lfdit  0   mi Lfdit i   pi Lm2 t i   qi grot i   1 Lfdit 1   2 Lm2 i 1 i 1 i 1   3 grot 1   t (3)To test for the existence of any long-run relationship we test whether all .1 The F-test has a non-standard distribution. The testing of thehypothesis is made by comparing the F-statistic with the upper and lowerbounds of critical values at 5 per cent level of significance. The appropriatecritical values for different number of regressors have been tabulated inPesaran and Pesaran (1997). There is evidence of a long-run relationshipbetween the variables if the F-test statistic exceeds their respective uppercritical values. On the other hand, cannot reject the null hypothesis of nocointegration if the test statistics less than the lower bound critical values. Thehypothesis remains inconclusive if the F-statistic lies between the upper andlower bound of the critical values and a decision will be made based on theECM version of the ARDL model. The ECM integrates the short run dynamicswith the long run equilibrium without losing long run information. 240
  • 7. Shahrudin, Yusof & Satar3.3 DataData spanning from 1970-2008 are obtained from various sources ofpublications. Data on FDI, rate of growth, openness, and governmentdevelopment expenditure are taken from various issues of Bank NegaraMalaysia reports. FDI is the direct investment in Malaysia under the Balanceof Payment account statement. In the case of Malaysia, FDI is measured asthe initial capital outlays in the approved projects by Malaysian IndustrialDevelopment Authority (MIDA). International Financial Statistics provides dataon exchange rate, money supply, and inflation rate. Statutory corporate taxdata is taken from the Department of Inland Revenue annual reports.4. FindingsTable 2 reports results on unit root test using Augmented Dickey Fuller (ADF)test. The finding suggests that all variables are integrated of order 1, I(1),except gro and infl which are I(0). This test suggests that an ARDL approachis the appropriate method of estimation since the variables are mixture of I(1)and I(0). Next we estimate equation (1) to determine the existence of anylong-run relationship between FDI and other macroeconomic variables. Themaximum order of lags in the ARDL model in both cases is n=1 due to theannual time series data utilized in the study. The calculated F-statistics forcointegration test is shown in Table 3 for equation (1) that includes all thevariables. The F-test has a non-standard distribution. The critical valuebounds for F-test are computed by Pesaran et al (1997). Table 2: Augmented Dickey-Fuller (ADF) Test for Unit Root Variable Levels First-differences Without With Trend Without Trend With Trend Trend Lfdi -1.346 -2.257 -5.448*** -5.418*** Lm2 -1.887 -2.456 -6.048*** -6.523*** Lgdp -1.703 -1.897 -5.025*** -5.164*** gro -4.927*** -5.298*** Ldev -1.994 -2.378 -4.899*** -4.887*** open -0.887 -1.750 -4.753*** -4.707*** Lexc -1.164 -2.905 -6.162*** -6.174*** tax 0.361 -1.853 -5.547*** -5.660*** infl -3.525** -3.932**Notes: i. ***, **, and * represent statistical significance at the 1%, 5%, and 10% level respectively. ii. The optimal lag lengths are determined by the Schwarz’s Information Criterion (SIC).The calculated F-statistic for F(Lfdi|Lm2, Lgdp, gro, Ldev, open, Lexc, tax, infl,crisis) is 1.1557 which is less than the lower bound critical value at 5 per centlevel of significance. This implies that the null hypothesis of no cointegrationcannot be rejected and therefore it suggests that there is no evidence ofcointegration among the variables in equation (1). This is not surprising asAng (2008) also faces the similar problem when some of the variables turn 241
  • 8. Shahrudin, Yusof & Satarout to be insignificant when in a model that includes all the variables. Theprocess of testing for the existence of a long-run relationship using boundtesting is repeated for other combination of variables by dropping some of thevariables in each of the equation. Finally, the presence of cointegration isdetected in F(Lfdi| Lm2, gro). The F-statistics of Wald test for F(Lfdi| Lm2,gro) is 4.8644 which exceeds the upper bound critical value at 5 per cent levelof significance. Therefore, the null hypothesis of no
  • Search
    Related Search
    We Need Your Support
    Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

    Thanks to everyone for your continued support.

    No, Thanks

    We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

    More details...

    Sign Now!

    We are very appreciated for your Prompt Action!